The present invention relates to a digital signal processing apparatus, and more particularly, to apparatus for processing compressed frequency components of a digital image signal to provide fade-in, fade-out and scene-change effects and further, to provide scrambling and descrambling of the digital image signal.
Audio-video equipment, such as a consumer VTR (video tape recorder) having a built-in camera (i.e. a camcorder), permit fade-in or fade-out of a video image during a scene change or the like. During this fade-in or fade-out operation, it is known to utilize a white fade for a white fade in/out and to utilize a black fade for a black fade in/out or the like.
FIG. 9 illustrates one example of an arrangement of a conventional fade-in and fade-out circuit. As shown in FIG. 9, a composite video signal is supplied to an input terminal 101. Since it is common that the composite video signal be transmitted at 75 ohms and 1Vp-p (peak to peak), the composite video signal is terminated by a terminating resistor 102 of 75 ohms and received by a buffer 103. An output from the buffer 103 is supplied through a capacitor 104 to a clamper 105. The clamper 105 is adapted to prevent a DC component of the pedestal level of the video signal from becoming unstable. The output of clamper 105 is supplied to a Y/C separating circuit 106 and a synchronizing separating circuit 107. The synchronizing separating circuit 107 extracts from the composite video signal a composite sync. signal from which a horizontal (H) sync. signal and a vertical (V) sync. signal are synthesized. The Y/C separating circuit 106 separates the composite video signal to provide a luminance signal Y and a chrominance signal C which are supplied to buffers 108, 109, respectively.
Typically the fade-in and fade-out operations are carried out by varying the resistance values of three variable resistors. Variable resistors 111 and 112 pass, or attenuate, the luminance signal Y and the chrominance sign C, respectively, thereby adjusting the levels of these signals. Another variable resistor 110 connects a power source voltage Vcc to a mixer 113, thus applying a DC voltage to the mixer. This voltage is referred to as a set-up voltage.
Normally, during a non-fade in or fade-out operation, the resistance value of variable resistor 110 is set to its maximum value and the resistance values of variable resistors 111, 112 are set to zero (short circuit) such that luminance signal Y passes through variable resistor 111, mixer 113 and buffer 114 and chrominance sign C passes through variable resistor 112 and buffer 115, whereby the original composite video signal is re-synthesized by mixer 116 and the output thereof is supplied to a buffer 118 through a switch 117. During a composite sync. period, switch 117 supplies the output of synchronizing separating circuit 107 to buffer 118. The output of buffer 118 is supplied to output terminal 120 through an impedance matching resistor 119.
The fade-out operation of the fade in/fade out circuit of FIG. 9 will now be described. The resistance values of variable resistors 111, 112 are progressively increased in response to a white fade-out command and simultaneously, the resistance value of variable resistor 110 is progressively decreased. When the resistance values of variable resistors 111, 112 are at their maximum and the resistance value of variable resistor 110 is at its minimum, luminance signal Y supplied to buffer 114 is held at the white level. A white fade-in operation is carried out by merely performing the reverse of the above described fade-out operation.
A black fade-out operation is performed when the resistance value of variable resistor 110 is set to its maximum value for the entire aforesaid fade-out operation. The black fade-in operation is the reverse operation of the black fade-out operation.
Satellite broadcasting companies and cable television companies which transmit scrambled video and audio signals via satellite or cable have long been commercially available. Television viewers cannot watch or listen to broadcast video and audio signals unless the broadcast signals are de-scrambled. In this instance, television viewers cannot enjoy picture and sound of such broadcasting unless a descrambler, leased by the broadcaster to the television viewer, descrambles the signal prior to its receipt by the television receiver. However, if video and audio signals are scrambled in such a fashion that the picture and sound are completely unrecognizable by the user, there is the risk that the user misunderstands this condition to be a malfunction of the TV receiver. But, if the user is able to recognize to some extent the sound accompanying the scrambled picture being broadcasted then the user may be interested in becoming a subscriber to that broadcast station. For this reason, video and audio signals are transmitted so that they can be recognized by the user to some extent. By way of example, there have heretofore been proposed various techniques such as changing lines in the vertical direction at every 1H (H being a horizontal scanning period), replacing each half of a line, carrying out the preceding operation at every several horizontal intervals and changing the processing at every several frames.
FIG. 10 shows an arrangement of a descrambler used in an existing cable TV box. An input terminal 41 in FIG. 10 receives signals supplied from a line which is normally supplied by a cable distributer to the viewer. The signal supplied to input terminal 41 is a scrambled RF (radio frequency) signal which is received by a tuner 43. An operation key 47 or a remote controller 42 are utilized to select the baseband signal which is to be supplied as an output by tuner 43. This baseband signal is supplied to a descrambler circuit 44 which descrambles the baseband signal. The descrambled signal is modulated onto a predetermined broadcast frequency, such as channel 3, by a modulating circuit 45 and then output from the cable box. This output signal is supplied to a tuner 46 contained within a TV receiver or VTR and tuned to the predetermined broadcast frequency to reproduce the video image and audio signal.
FIG. 11 shows an arrangement of a descrambler utilized with the existing satellite broadcast systems. As shown in FIG. 11, an RF signal from a BS/CS antenna 51 is supplied to a BS/CS tuner 52 provided within a VTR or TV. Channel selection is carried out by utilizing an operation key 58 or a remote controller 57. A detected output signal 53 from tuner 52 is supplied to a descrambler 54. A video signal 55 and an audio signal 56 thus descrambled by the descrambler 54 are each output twice, for example. One set of video and audio signals 55, 56 are supplied to the line input of the VTR, and the other set of video and audio signals 55, 56 are supplied to the line input of the TV receiver.
A method of releasing the scrambler, i.e., descrambling, and a method of collecting by the program distributer service charges will be described next. Descrambling is carried out by using an ID (identification) number assigned to the cable box (descrambler apparatus) and a key code contained in a corresponding broadcast signal. When a subscriber (television viewer) pays a service charge, the broadcasting enterprise transmits the key code recognized by the cable box to descramble the broadcast signal. The key code and the ID number are related by enciphering.
There are many pay channel broadcasting systems and they tend to increase. In accordance therewith, the number of descramblers are used by the user (television viewer) increases as the number of subscriptions increases. Mainly in European countries, there is available a descrambler that resembles an IC telephone card, known as a "smart card" that is commonly used for many types of descramblers. Therefore, the user can receive several pay channel broadcasting programs using one descrambler.
There have heretofore been proposed audio-video equipment such as a digital broadcasting system or a teleconference system in which a video signal is digitized and transmitted via radio waves or cable and a digital VTR records the digital signal on a magnetic tape. When a video signal is digitized, the amount of information representing the video image is quite large and, therefore, a high-efficiency coding system that compresses data as much as possible is frequently utilized. Of the various high-efficiency coding systems, discrete cosine transformation (DCT) is commonly used.
During discrete cosine transformation, an image of one frame is formed into, for example, an (8.times.8) block structure, where each block (of pixels) is processed by utilizing a discrete cosine transform, which is one type of an orthogonal transform. The transformed data is further processed, as by weighting, requantization or the like. According to the above-described process, energy components within each (8.times.8) block can be concentrated into a certain area, and data in other areas are converted to "0" or negligible values close to "0". If the later data are removed, the amount of data needed to represent the image can be reduced to some extent.
Despite the above-mentioned process, the amount of data is still large. Therefore, the coefficient data resulting from the discrete cosine transformation are processed using a Huffman coding process in which the coefficient data are converted into data having different bit lengths in response to the probability at which a signal is generated. Such processing may include variable length coding. Hence, data can be compressed to the extent needed for a consumer digital VTR to record and reproduce digitized video pictures. In order to facilitate data processing during reproduction, framing is carried out in which a coded signal is inserted into a data area within a sync. block of a constant length and the sync. block is formed by adding a parity, a synchronizing signal, an ID signal or the like for protecting such data.
A two-dimensional Huffman coding system is frequently utilized as one of the variable length coding systems because of its high efficiency. This coding system is referred to as an amplitude run length coding system in which one event is defined to a value of each coefficient whose value is not "0". If an (8.times.8) block, for example, is processed by the two-dimensional Huffman coding system, then it is frequently observed that all of the data equals "0" after a certain amount has been processed. In this instance, no coefficients which have a value equal to "0" are supplied and instead, a code EOB (End of Block) indicative of the end of the block is inserted.
In a digital VTR using a magnetic tape and in a digital disc recording and reproducing apparatus using a disc-shaped recording medium, it is customary that video data of one or a plurality of field or frames are recorded on a plurality of tracks. In this instance, if the number of tracks per field/frame are different for each field/frame, there is then the disadvantage that processing is not satisfactorily effected. Therefore, a constant number of tracks per field (or frame) is frequently selected. Although the amount of data remains constant in a system in which data is not compressed, if data is compressed using the above-described DCT process, the amount of data varies depending upon the picture content. Further, because data is processed by a variable length coding system, the amount of data in a predetermined period is constantly changing. In order to record such data in a constant number of tracks, it is necessary to have a buffering system which makes the amount of data in a predetermined period less than some predetermined target value.
By way of example, there has heretofore been proposed a buffering system in which a resultant amount of data is made less than a target value even for a period of one field or one frame by controlling the amount of data in a predetermined period (referred to as a buffering unit) which is shorter than one field or one frame. According to this buffering system, data which is to be transmitted can be made less than a target value by re-quantizing the coefficient data of the DCT transformed AC component of the signal with proper quantization step. A quantization step code or quantization number indicative of the quantization step is transmitted with the coded data.
FIG. 6 shows an example of a sync. block of component signal transmission data according to the prior art. As shown one buffering unit is formed of 5 sync. blocks. A block synchronizing signal SYNC is located at the starting position of each sync. block, followed by an ID signal, a quantization number QNO, and auxiliary information AUX. DCT coefficients are disposed in the data area of each sync. block followed by parity data (error correction code). It is common for the error correction code to be Reed Solomon code. Data which is coded by a variable length coding system is arranged (this process being referred to as "packing") into 8-bit data and then recorded in the data area. In general, because a buffering unit has only one quantization number QNO, the quantization numbers QNO of the respective sync. blocks have the same value.
The data area will now be described with reference to FIG. 7. In the case of an (8.times.8) block, for example, if sample data of 8 pixels.times.8 lines, shown in FIG. 7A, are processed in a DCT (discrete cosine transform) fashion, then such data is converted into coefficient data formed of a direct current component DC and alternating current components AC1 to AC63 as shown in FIG. 7B. Arrows in FIG. 7B represent zigzag scannings that are generally carried out. The direct current component DC expresses an average luminance value on the basis of a DCT definition and it is known that the direct current component DC has a value approximately twice the average of the absolute values of the 64 pixels. In addition, the direct current component DC represents the maximum energy of the block and therefore is the most important component during transmission.
When data is processed by a variable length coding system, the number of bits of one word can be detected by sequentially checking the input data. Accordingly, if an error occurs even in one place, the word interval cannot be detected and as a result, succeeding data cannot be checked, causing a propagation error. That is to say, although the variable length coding system is excellent for efficient data compression, it is very weak against errors. Therefore, it is customary that the direct current component DC is not processed by the variable length coding system, and only the alternating current components AC1 to AC63 are processed by the variable length coding system.
In FIG. 6, a 9-bit DC component, a 1-bit motion flag M and a 2-bit activity code (definition of picture) are typically arranged as the most important words. Thereafter, alternating components in the DCT block that had been processed by the variable length coding system are sequentially filled in the direction of the alternating current components AC1 to AC63. In this example, Y has a fixed area AC-L of 12 bytes, C has a fixed area AC-H of 6 bytes and the alternating current components are stored therein as described above. At that time, when the alternating current components are stored up to the EOB of the DCT block within the fixed AC-L area, remaining bits in the fixed AC-L area are newly defined as variable AC-H areas. Then, the variable AC-H area and the fixed AC-H area are combined as a new AC-H area, and components that had not yet been stored in the fixed AC-H area are stored therein sequentially. As a method of storing components in the data area, in addition to the above-mentioned regular framing method, there is known a front framing method of sequentially storing all data or the like.
When the above-mentioned fade-in and fade-out effects are achieved by digital equipment such as a digital VTR or the like, digital data of the luminance signal Y and the chrominance signal C are reconverted into analog data and processed by the circuit arrangement shown in the latter half of FIG. 9. It is generally known that, if an analog circuit and a digital circuit are provided as a mixed or hybrid circuit, then noise generated by the digital circuit or the like enters and interferes with the analog circuit. For example, problems such as ground pattern layout or the like may be introduced. For this reason, it is preferable to reduce the analog sections as much as possible.
Another problem with the above-mentioned circuit is that the fade-in and fade-out effects can only be achieved in the analog output of the digital VTR but cannot be achieved in the digital output. That is to say, users can enjoy the fade-in and fade-out effects on the existing TV monitor receiving the analog signal output, but cannot enjoy such effects on a future TV monitor which receives a digital signal directly. Audio-video equipment becomes expensive when fade-in and fade-out are achieved by using special circuitry in the digital output.
Typically, a scrambler may be implemented by a wide variety of circuits such as an analog-to-digital (A/D) converter for converting an analog signal into a digital signal, one or a plurality of frame or field memories and line memories for storing the converted digital signal, a digital-to-analog (D/A) converter for reconverting a digital signal into an analog signal, a control circuit for controlling the above-mentioned circuit elements or the like.
The scrambler utilized by a broadcasting enterprise is usually of a professional type and only one scrambler need be used for the transmission of TV signals. Ordinarily, the size and cost of the circuitry are not objectionable. However, a user's descrambler requires circuits similar to those described above, and here such a descrambler becomes expensive.
On the other hand, a descrambler utilized with a digital TV broadcasting system, will not require an A/D converter, D/A converter and so on. Further, almost all of the digital circuits within the descrambler can be commonly used by a digital tuner. That is, when the present broadcast/receiver system, i.e., antenna--tuner--descrambler--TV is changed to the future system, i.e., antenna--digital tuner--TV, it is expected that the digital tuner will include the descrambler function. In FIG. 11, the tuner 52 and the descrambler 54 will be integrated as the digital tuner. At that time, the key for descrambling may be supplied in the form of a card such as an IC card, magnetic card or the like, supplied by the broadcast enterprise.